The major objective is to identify experimental approaches that will reduce the rate of tumor cell dissemination in spontaneous metastasis models of neuroectodermal tumors. Major hypotheses to be tested are that either immunotherapy modalities designed to activate immune effector cells and/or manipulations that interfere with extracellular matrix (ECM) degradation by metalloproteinases (MMPs) are adequate to disturb the multi-step metastasis process sufficiently to decrease the rate of tumor cell invasion and metastasis in vitro and in vivo. Encouraging initial clinical results with anti-GD2 monoclonal antibody (mAb) in neuroblastoma provided the rationale to focus on the design and evaluation of novel immunotherapy modalities for this malignancy, including recombinant mAb fusion proteins with cytokines IL-2, TNF- alpha, TNF-beta, GM-CSF. The potential of these modalities will first be tested in a syngeneic spontaneous metastasis model for murine neuroblastoma in A/J mice and then in such a model for human neuroblastoma in scid mice engrafted with human immune effector cells. Comparisons will be made between the respective efficacy of killing metastatic melanoma cells in vitro and in vivo with either HL-A Class I- restricted, melanoma-specific allogeneic cytotoxic T cells or recombinant bispecific constructs between anti-tumor mAbs and mAbs directed to either CD3 or FcgammaIII receptor. Attempts to decrease the invasive/metastatic potential of melanoma cells by manipulating MMP- derived ECM degradation will include: 1) transfection of melanoma cells with cDNA encoding the natural tissue inhibitor-2 of MMP, and 2) targeting of melanoma metastasis with a mAb-fusion protein containing a procollagenase peptide that inhibits tumor cell invasion in vitro. Efforts to establish which MMPs can render non- or poorly metastatic melanoma more aggressively metastatic include: 1) selective neutralization of individual MMP gene products by anti-sense RNA, and 2) up-regulation of individual MMPs by transfection of tumor cells with cDNAs encoding them. These studies will be rounded out by identifying selected integrins as potential regulatory elements of MMP expression and selecting those that can modulate poorly metastatic tumors to metastasize more aggressively. It is anticipated that some of the results from our studies may ultimately contribute to improve the treatment of metastatic cancer.